Inconsistency in aortic stenosis severity between CT and echocardiography: prevalence and insights into mechanistic differences using computational fluid dynamics

Figures

• 

  Figure 1. Flow diagram showing recruitment of patients in the study. AV, aortic valve; CTCA, CT cardiac angiogram; TAVI, transcatheter aortic valve replacement; TTE, transthoracic echocardiogram.

• 

  Figure 2. Measurement of CT aortic valve (AV) area. CT cardiac angiography images demonstrating a method of obtaining AV area in a patient with tricuspid valve with calcification. Midsystolic phase of the cardiac cycle with the maximum opening of the AV is first selected. Starting from a coronal image (A) with non-coronary (left) and left leaflets (right) (black arrows), one of the crosshair is aligned (B) such that it passes parallel to the AV plane and the other perpendicular through the AV orifice, resulting in a left ventricular outflow tract view (C). The crosshair is further aligned parallel to the valve plane (D) to produce a true short-axis view through the AV (E). By scrolling in this plane, the smallest orifice is obtained and direct planimetry of the orifice (F) is performed along the inner edges of the leaflets to obtain the AV area.

• 

  Figure 3. Cases demonstrating the correlation of CT aortic valve area (CT-AVA) with computational flow dynamics. All cases have echocardiographic Vmax>4 m/s and mean gradient>40 mm Hg. (A) Upper row demonstrates CT-AVA by direct planimetry, while lower row (B) depicts computational fluid dynamics simulation (red colour representing highest velocity, blue intermediate and yellow lowest). Case 1 represents consistent aortic stenosis grading with CT-AVA=0.8 cm2, flow rate=517 and flat velocity profile of 1.1. Cases 2, 3 and 4 represent inconsistent grading with CT-AVA=1.7 cm2, 1.5 cm2 and 1.2 cm2; high flow rate of 604, 701 and 530; and skewed velocity profile of 1.3, 1.24 and 1.3, respectively.